[Enhancement]: full context controllable per session

[stdweird]

What features would you like to see added?

the environment variable RAG_USE_FULL_CONTEXT is too coarse, and it could be be better to provide some switch to enable per session

as an extra, showing an estimation in tokens size of each document in the upload management to help decide to turn it on or off

More details

when using uploaded documents, users can decide if it is possible or wanted to use the full document as context. criteria can be size of the document (5 page pdf vs 500 page pdf), and ofcourse impact on cost.

showing the total size of a document in tokens (and perhaps also characters) might help user to estimate the feasibility (context window and budget). ofcourse, most providers have their own tokenizers; so would need to fix on whatever rag_api can deal with, and thus make it clear to the users that this is an estimate rather than an exact number.

Which components are impacted by your request?

No response

Pictures

No response

Code of Conduct

• I agree to follow this project's Code of Conduct